Non-Stationary Texture Synthesis by Adversarial Expansion

Peng Liu June 3, 2018

Abstract

The real world exhibits an abundance of non-stationary textures. Examples include textures with large-scale structures, as well as spatially variant and inhomogeneous textures. While existing example-based texture synthesis methods can cope well with stationary textures, non-stationary textures still pose a considerable challenge, which remains unresolved. In this paper, we propose a new approach for example-based non-stationary texture synthesis. Our approach uses a generative adversarial network (GAN), trained to double the spatial extent of texture blocks extracted from a specific texture exemplary. Once trained, the fully convolutional generator is able to expand the size of the entire exemplar, as well as of any of its sub-blocks. We demonstrate that this conceptually simple approach is highly effective for capturing large-scale structures, as well as other non-stationary attributes of the input exemplary. As a result, it can cope with challenging textures, which, to our knowledge, no other existing method can handle.

CCS Concepts: • Computing methodologies → Appearance and texture representations; Image manipulation; Texturing;

YANG ZHOU, Shenzhen University and Huazhong University of Science & Technology ZHEN ZHU and XIANG BAI, Huazhong University of Science and Technology DANI LISCHINSKI, The Hebrew University of Jerusalem DANIEL COHEN-OR, Shenzhen University and Tel Aviv University HUI HUANG, Shenzhen University

https://arxiv.org/pdf/1805.04487.pdf

This work is the most amazing one that I have never seen. I have some minds about it:

How about to use it do microscopy image synthesis?
It is definitely able to create customized production decorated with paint arts.
Something else? see this video

A mixed-scale dense convolutional neural network for image analysis

Peng Liu June 3, 2018

Deep convolutional neural networks have been successfully applied to many image-processing problems in recent works. Popular network architectures often add additional operations and connections to the standard architecture to enable training deeper networks. To achieve accurate results in practice, a large number of trainable parameters are often required. Here, we introduce a network architecture based on using dilated convolutions to capture features at different image scales and densely connecting all feature maps with each other. The resulting architecture is able to achieve accurate results with relatively few parameters and consists of a single set of operations, making it easier to implement, train, and apply in practice, and automatically adapts to different problems. We compare results of the proposed network architecture with popular existing architectures for several segmentation problems, showing that the proposed architecture is able to achieve accurate results with fewer parameters, with a reduced risk of overfitting the training data.

Danie ̈lM.Pelt and James.Sethian

https://slidecam-camera.lbl.gov/static/asset/PNAS.pdf

This work uses the dense connection and multiple factors dilated convolution to extract the mixed-scaled features in the same layer.

The idea may have some advantages of extracting mixed scale features. However, it does not provide more comparison study to give us confidence about the effectiveness and novelty of this work.

What’s “self-motivated”

Peng Liu June 2, 2018

10 Traits of Self-Motivated People

Highly naturally-motivated people …

1. Have extraordinary purpose

They set their goals high and are always reaching for the top.

2. Are willing to take risks

They have the confidence to step out of their comfort zone to try new ideas or strategies.

3. Participate fully in life

They take part in their organizations, in their families and in their communities.

4. Are energetic

They’re willing to jump in and take things on — and have the mental energy necessary to get it done.

5. Are humble

They’re not afraid to admit they don’t know it all. They’re eager to keep growing.

6. Are committed to lifelong learning

They know their stuff and are always in the process of learning more.

7. Possess an attitude of success

They act as if it were impossible for them to fail, as if their success is a done deal.

8. Are persistent — with options

They never give up until they succeed, but they’re willing to try a variety of options to get what they want rather than to keep hitting their head against the same brick wall.

9. Strive for health in all aspects of their lives

They work hard to take care of themselves physically, emotionally, spiritually, intellectually and socially.

10. Rise above adversity

They rise above the small stuff to achieve greatness in their lives.

Find yourself lacking in some of these areas? Identify others you know who possess these traits and ask them to mentor you or find out how they went about developing these attributes, Cohen recommends. Read about famous people who have the characteristics you’d like to improve in yourself and see what you can learn from their experiences.

Make a concerted effort to integrate the traits of highly motivated people into your behaviors, and you’ll be amazed at how much more you can achieve in your personal and professional life, Cohen concludes. [Source]